DESCRIPTION: This is a proposal for a program of research to develop a theoretical framework for characterizing situational influences on strategy use and the mechanisms that may be responsible for developmental changes in cognitive strategies in children with and without mental retardation. This framework will be based on a set of neural network models and on behavioral experiments with school-aged children designed to test implications of the models. The interaction between empirical research and neural network models is designed to suggest new psychological mechanisms and biologically plausible processes that may account for developmental changes in strategy use in children. Three experiments will investigate situational influences on the development of external memory strategies with groups of children with and without mental retardation. Two neural network models will be modified or adapted to simulate the conditions of each experiment. The goodness-of-fit between benchmark measures obtained from the empirical results and the simulated strategy and accuracy output from each model will be evaluated, and the role of the strategy mechanisms of each model will be analyzed.